Minireview: G Protein-Coupled Estrogen Receptor-1, GPER-1: Its Mechanism of Action and Role in Female Reproductive Cancer, Renal and Vascular Physiology

Abstract

Using cDNA cloning strategies commonly employed for G protein-coupled receptors (GPCR), GPCR-30 (GPR30), was isolated from mammalian cells before knowledge of its cognate ligand. GPR30 is evolutionarily conserved throughout the vertebrates. A broad literature suggests that GPR30 is a Gs-coupled heptahelical transmembrane receptor that promotes specific binding of naturally occurring and man-made estrogens but not cortisol, progesterone, or testosterone. Its “pregenomic” signaling actions are manifested by plasma membrane-associated actions familiar to GPCR, namely, stimulation of adenylyl cyclase and Gβγ-subunit protein-dependent release of membrane-tethered heparan bound epidermal growth factor. These facts regarding its mechanism of action have led to the formal renaming of this receptor to its current functional designate, G protein-coupled estrogen receptor (ER) (GPER)-1. Further insight regarding its biochemical action and physiological functions in vertebrates is derived from receptor knockdown studies and the use of selective agonists/antagonists that discriminate GPER-1 from the nuclear steroid hormone receptors, ERα and ERβ. GPER-1-selective agents have linked GPER-1 to physiological and pathological events regulated by estrogen action, including, but not limited to, the central nervous, immune, renal, reproductive, and cardiovascular systems. Moreover, immunohistochemical studies have shown a positive association between GPER-1 expression and progression of female reproductive cancer, a relationship that is diametrically opposed from ER. Unlike ER knockout mice, GPER-1 knockout mice are fertile and show no overt reproductive anomalies. However, they do exhibit thymic atrophy, impaired glucose tolerance, and altered bone growth. Here, we discuss the role of GPER-1 in female reproductive cancers as well as renal and vascular physiology.

Estrogen promotes a multitude of biochemical actions that are measured with distinctly different kinetics. Categorically, these signaling events are described as rapid or “pregenomic” events (second messenger and protein/lipid kinase activation) that occur within minutes of estrogen exposure and delayed or genomic transcriptional responses that are most conveniently measured using an hourly metric. In this context, it is important to recognize that pregenomic actions ultimately may lead to gene transactivation.

It is universally appreciated that estrogen receptors (ER), ERα and ERβ, which belong to the nuclear steroid hormone receptor superfamily, function as hormone-inducible transcription factors and induce estrogen-dependent gene transactivation. However, the physical identity and nature of the receptor(s) that manifest pregenomic estrogen action have been a matter of healthy debate. One of the most promising candidate receptors to fulfill this role is the former orphan receptor, G protein-coupled receptor (GPCR)-30 (GPR30), more recently referred to by its proposed functional designate, G protein-coupled ER (GPER)-1. The signaling mechanisms employed by GPER-1 that allow for stimulation of adenylyl cyclase and release of membrane-tethered epidermal growth factor (EGF)-like polypeptides are not particular to GPER-1 and are familiar to many other GPCR and have been reviewed previously (1, 2). Alternative models of pregenomic estrogen action have been suggested and involve intact ER protein, or derivatives of, as well as the possibility that ER and GPER-1 may act coordinately for this purpose (3). However, there are many unresolved issues regarding the molecular mechanisms by which nuclear steroid hormone receptor recruit, and activate, signaling effectors necessary for activation of the plasma membrane-associated enzymes that manifest pregenomic estrogen signaling.

Numerous studies demonstrating estrogen pregenomic signaling in GPER-1-positive, ER-negative cells indicate that GPER-1 can act as a “stand alone” receptor. In addition to the fact that ER and GPER-1 are linked to different signaling mechanisms in reproductive cancers, their actions are independent by several measures, including the following facts: 1) independent expression of ER and GPER-1 is observed in breast tumors and in cultured breast cancer cells lines (4, 5); 2) ER and GPER-1 display different binding affinities for various estrogens, phytoestrogens, and xenoestrogens and are differentially activated by them (5–7); 3) ER antagonists serve as GPER-1 agonists (2); 4) distinct functions of GPER have been identified using selective GPER-1 agonists and antagonists (8); 5) ER and GPER null mice exhibit distinct phenotypes (9–11); and finally 6) GPER-1 and ER differentially associate with markers of female reproductive cancers (2, 12, 13).

GPR30 As a GPCR for Estrogen

Unlike ER, which was isolated by a classical protein chemistry strategy well suited for a soluble receptor, GPR30 was discovered by molecular cloning approaches that have been widely successful for identifying a large number of GPCR. In the late 1990s, no less than six different laboratories employing independent molecular cloning strategies reported the isolation and preliminary characterization of a GPCR homologue, which was assigned the orphan designation, GPR30, because the identity of its cognate ligand was unknown (14–19). Based upon its structural homology to angiotensin II receptors, and other chemotactic peptides receptors, it was presumed that the ligand for GPR30 was a peptide. Despite this homology, an assortment of chemotactic peptides, including IL-8, growth-regulated protein alpha; monocyte chemotactic protein-1; monocyte chemotactic protein-3; macrophage inhibitory protein-1a; complement factors 3a and 5a; regulated upon activation, normal, T-cell expressed and secreted; and lymphotoxin beta-4, and other peptide ligands, such as angiotensin II and angiotensin IV, was screened and shown not to bind GPR30 (15, 16).

The experiments that led to the discovery that GPR30/GPER-1 is a Gs-coupled receptor that promotes estrogen-dependent activation of adenylyl cyclase and epidermal growth factor receptor (EGFR) transactivation have been reviewed in depth elsewhere (2, 20). Its mechanism of action is not all that different among GPCR, although our new evidence suggests that GPER-1 employs a unique mechanism of receptor down-modulation and desensitization (discussed below).

Receptor Binding Characteristics of GPER-1

Specific estrogen binding activity for ER is readily measured by radioreceptor assays using intrinsically labeled ³H-17β-estradiol as tracer. Employing this standard approach, ER binding affinities for 17β-estradiol are typically in the subnanomolar range [dissociation constant (K_d) = 0.1–1.0 nm] using detergent-free cellular homogenates prepared from a tissue source that abundantly expresses ER (7). Measurement of specific estrogen binding to a membrane-associated ER (ER or GPER-1) is a more difficult challenge due to the relatively high background binding of this lipophilic ligand to the lipid-rich plasma membrane preparations and their relatively low levels of expression (21). For most GPCR, including GPER-1, only a small proportion of the receptor is expressed on the cell surface. Therefore, stable expression of novel steroid receptors, such as GPER-1 and membrane progestin receptors, and selection of clones with high receptor expression by limiting dilution or by other techniques, such as cell sorting, are often required to express sufficient amounts of the membrane protein for characterization of receptor functions (21, 22). Unlike the nuclear steroid receptors, GPER-1 and membrane progestin receptors are very labile, so ligand-binding assays with short incubation times at 4 C in the presence of protease inhibitors are used to limit degradation.

The membrane filtration assay is the standard method to separate ligand bound to plasma membrane fractions from free ligand and has been widely used to measure receptor binding of a broad range of ligands (23). Specific estrogen binding to plasma membranes of ER-negative cells expressing GPER-1 was first demonstrated for the human wild-type and recombinant receptor in SKBR3 and HEK293 cells, respectively, using this procedure and subsequently for GPER in a teleost species, Atlantic croaker (6, 24). Human and croaker GPER-1 were shown to have high affinity, limited capacity, displaceable, single binding sites for 17β-estradiol characteristic of steroid receptors. The same protocol was used by another research group to demonstrate [³H]-17β-estradiol binding to zebrafish GPER-1 (25). The 17β-estradiol binding affinities of the recombinant human (K_d = 3.3 nm), croaker (K_d = 2.7 nm), and zebrafish (K_d = 2.3 nm) GPER-1 are very similar and considerably lower than that of ER (6, 24–26). The rates of association and dissociation of [³H]-17β-estradiol binding to these GPER are very rapid and are completed within a few minutes, which is typical of steroid membrane receptors (6, 24, 25). The ligand binding of these GPER is specific for estrogens, because testosterone, cortisol, and progesterone have very low binding affinities for the receptor, approximately 0.1% that of 17β-estradiol (6, 24). As predicted, treatments that uncouple G proteins from GPCR reduce their ligand binding affinities and cause decreases in [³H]-17β-estradiol binding to both human and croaker GPER (6, 24). The finding that only minor amounts of specific [³H]-17β-estradiol binding can be detected in the microsomal fractions of SKBR3 cells, which contain large amounts of GPER-1, is likely due, in part, to the lack of GPER-1 coupling to G proteins in this intracellular compartment (27). Specific estrogen binding to GPER-1 has also been demonstrated by a third research group using an estrogen derivative conjugated at the 17 position to Alexa Fluor 546 to create a large fluorescent molecule (5). In contrast to the results of receptor binding assays using the [³H]-17β-estradiol ligand, estrogen-Alexa Fluor 546 binding was only detected in the perinuclear compartment of COS-7 cells transiently transfected with a human GPR30-green fluorescent protein(GFP) chimera, where the recombinant receptor protein had been localized by immunofluorescent analysis (5). The binding affinity of 17β-estradiol calculated from this heterologous competition assay (K_i, inhibitory dissociation constant) was approximately half (K_i = 6 nm) (5) of that determined from saturation analysis (K_d) of [³H]-17β-estradiol binding to cell membranes (6, 24, 25), which is consistent with a lower 17β-estradiol binding affinity for GPER-1 intracellularly, where it is not coupled to heterotrimeric G proteins. Finally, a fourth group of investigators confirmed the presence of estrogen binding in SKBR3 cells in a homologous competition assay using [³H]-17β-estradiol (28).

Despite the independent demonstration by these researchers that GPER-1 binds estrogen, the proposed function of GPER-1 as a membrane ER has been challenged by several other research groups on the basis of their negative results (3, 29–31). One of these studies employed a nonstandard radioreceptor assay to measure 17β-estradiol binding to whole COS-7 cells transiently transfected with GPER-1 (29). Unlike the membrane filtration assay, there are no reports to our knowledge of the successful application of a whole-cell binding assays to detect specific binding of steroids to membrane receptors. This is not surprising, because whole-cell assays are inappropriate for measuring membrane binding of lipophilic ligands such steroids, which readily enter intracellular compartments from which the potentially large pool of unbound radiolabeled steroid cannot be removed by washing (21). As predicted, no differences in total and nonspecific [³H]-17β-estradiol binding to cells stably transfected with GPER-1 were detected in the whole-cell assay, whereas specific [³H]-17β-estradiol binding could be detected in plasma membrane fractions isolated from the same batch of cells in the membrane filtration assay (21). In another study, plasma membrane fractions were incubated with a low concentration (1 nm) of [³H]-17β-estradiol at 37 C for 1 h (30), conditions suitable for the detection of membrane-bound ER but not for GPER-1, because they would result in incomplete saturation of GPER-1 binding sites and increased degradation of the receptor. Not surprisingly, no significant [³H]-17β-estradiol binding to plasma membranes of SKBR3 cells was detected with this incubation procedure in a membrane filtration assay, whereas large amounts of binding were detected in ER-positive MCF7 cells (30). Recently, an N-terminal truncated variant of ERα, ERα-36, has been detected in SKBR3 cells and in GPER-1-transfected HEK293 cells, and evidence was presented that estrogen signaling in these cells is not mediated through GPER-1 but instead occurs through ERα36 (31). However, estrogen binding and signaling through GPER-1 also occurs in fish and human cells that lack ERα36 or any other truncated forms of the ER, indicating that GPER-1 can function alone as a membrane ER (21, 24, 32).

Human and croaker GPER display very similar estrogen, phytoestrogen, and xenoestrogen binding specificities in competitive binding assays (5, 24), which are similar overall to those of the nuclear ER in these species but show some distinct differences. For example, diethylstilbestrol, which has binding affinities similar to, or greater than, 17β-estradiol for the ER (26, 33), shows very low affinities for both human and croaker GPER (6, 24). On the other hand, the binding affinity of bisphenol A to human GPER is 8–50 times higher than its affinity to the ER (34). Although the ER antagonists ICI 182,780 and tamoxifen can also compete for [³H]-17β-estradiol binding to human and croaker GPER-1, they exert opposite actions, mimicking 17β-estradiol's actions in both wild-type and recombinant human and fish models (5, 24). Both human and fish GPER are coupled to stimulatory G proteins and act through similar signaling pathways to activate adenylyl cyclase and transactivate EGFR (1, 5, 24, 35). The finding that the estrogen signaling functions of GPER are remarkably similar in mammals and fish, which diverged from the vertebrate lineage over 200 million years ago, suggests that its physiological role as a membrane ER is a basic, conserved function in vertebrates.

Cell Biological Functions: Lessons Learned from Genetically Deleted Mice, Human Reproductive Cancer Biopsy Tissue, and Renal and Vascular Physiology

Synthetic selective compounds for GPER-1 and genetic approaches that manipulate GPER-1 expression have proven useful for discerning its physiological and pathophysiological roles. Excellent reviews have been written about the use of GPER-1-selective ligands, including the GPER-1 agonist G-1 and the GPER-1 antagonist G-15, to investigate disease models (8, 36). Studies examining the cell biological action of GPER-1 have revealed that it promotes modest mitogenic responses as well as homeostatic responses that are typically attributed to GPCR. Stimulation of GPER-1, by either 17β-estradiol or the selective GPER-1 agonist, G-1, has been linked to proliferative responses in cancer cells derived from reproductive tissue, including breast (37), endometrium (38, 39), ovary (40), and testis (41), and also from thyroid tissue (42). Similarly, GPER-1-mediated proliferative responses have been measured in breast cancer cells in response to the phytoestrogen, genistein (43), and in spermatogonial (44) and seminoma (45) cells after exposure to the xenoestrogen, bisphenol A. However, in some experimental settings, GPER-1 stimulation inhibits proliferation (46). It is interesting to note that estrogen stimulation of human SKBR3 breast cancer cells that do not express ERα or ERβ but express GPER-1 and elevated levels of erbB2 induces pro-HB-EGF release and prolonged erbB1 tyrosyl phosphorylation and downstream erk-1/-2 activation (47) but only modest mitogenic effects (29). However, SKBR3 cells undergo a robust mitogenic response when treated with low doses of recombinant HB-EGF (1 ng/ml) (Filardo, E. J., and J. A. Quinn, unpublished observation), suggesting that 17β-estradiol may not induce a strong proliferative response in these cells due to partial occupancy of erbB1 by locally released HB-EGF.

Although the EGFR is an integration point for signaling mediated by GPCR, the transmembrane signaling events that lie between the GPCR and the EGFR and connect the activation of nonreceptor intracellular Src family kinases with exoplasmic release of membrane-tethered EGF-like polypeptides have remained unclear. Recent work has implicated integrin α5β1, the primary fibronectin receptor in epithelia, as a necessary transmembrane intermediary for GPER-1-mediated EGFR transactivation (Fig. 1) (48). These data show that GPER-1 action in breast cancer cells promotes Src-like kinase-dependent activation of integrin α5β1 as measured by its association with the signaling adapter Shc, and the recruitment of ligand-occupied integrin α5β1 conformers to fibrillar adhesions, specialized adhesion sites that are involved in the formation of fibrillar fibronectin and the assembly of a fibronectin matrix. Furthermore, we show using genetic deletion models, inhibitory antibodies, and by soluble Arginine-Glycine-Aspartic acid peptides that integrin α5β1 engagement is necessary for EGFR transactivation. These data expand upon our earlier work that showed that GPER-1-mediated Gβγ- protein subunit signaling triggers the release of intracellular calcium (49) and that Src-like kinases (47) are required for EGFR transactivation. Thus, collectively, these data support the idea that GPER-1 coordinates the release of EGF-like polypeptides and the formation of a provisional fibronectin matrix, activities that likely would allow for cellular survival during homeostasis of glandular tissue and possibly for infiltrating cancer cells that had escaped the confines of the glandular epithelium. Moreover, fibronectin matrix assembly is required for anchorage-independent growth, the best predictor for experimental metastases. It is interesting to note that successful implantation of mammary tumor xenografts is facilitated by the coadministration of exogenous fibronectin (50), suggesting a survival advantage to tumor cells that interact with fibronectin. The finding that GPER-1 promotes integrin activation is consistent with previous descriptions that have shown that GPCR agonists are often connected to integrin activation by “inside-out” signaling in leukocytes and platelets (51, 52).

Fig. 1.

GPER-1-mediated EGFR transactivation. GPER-1 employs a convoluted intracellular signaling scheme that is a recurrent paradigm for transactivation of the EGFR and used by many members of the GPCR superfamily. GPER-1 stimulation with 17β-estradiol induces Gβγ-subunit protein-mediated release of intracellular calcium from endoplasmic reticulum stores resulting in the activation of Src-like nonreceptor tyrosine kinases. This action causes the clustering and activation of integrin α5β1, promoting its concentration in focal and fibrillar adhesions and inducing matrix metalloproteinase-dependent release of membrane tethered pro-HB-EGF and subsequent activation of the EGFR. In this manner, estrogen action via GPER-1 coordinates the release of local EGF-related polypeptides and induces fibronectin matrix assembly, cellular activities that are associated with cellular survival. AC, Adenylyl cyclase; MMP, matrix metalloproteinase; Fn, fibronectin; E, estrogen.

Despite the fact that GPER-1 promotes cell biological responses in reproductive tissue, GPER-1-deleted mice do not show the gross reproductive anomalies that are expressed in ER knockout mice (it should be noted that although ERα-deleted mice are infertile, this is not the case for ERβ mice) (9–11). Instead, GPER-1 knockout mice express a phenotype that shows a number of less overt traits, including thymic atrophy, impaired glucose tolerance, and altered bone growth. Several explanations can be provided for the failure to observe a reproductive phenotype, including 1) gene redundancy, 2) incomplete penetrance suggesting a requirement for trans-acting factors (or physiological or pathophysiological conditions), or 3) compensation by unrelated gene products, including ER or other gene products. For reasons mentioned above, it is possible that a major role of GPER-1 in the reproductive system is to promote cellular survival by estrogens during homeostatic maintenance of the mammary gland. These same cellular mechanisms may allow for the survival of mammary adenocarcinoma cells that find themselves outside the context of the mammary gland whether by acquired invasive potential or as the result of a breakdown in the integrity of the basement membrane of the gland during normal remodeling of mammary epithelia.

The best evidence showing that GPER-1 acts independently from ER in reproductive cancers comes from studies of its expression in biopsy specimens from patients with reproductive tumors (4). GPER-1 expression occurs independently of nuclear ER, ERα, and ERβ as evidenced in primary tumors and their isolates. Moreover, GPER-1 and ER display distinct patterns of association with clinicopathological parameters that predict disease progression of reproductive cancers. Namely, in breast cancer, GPER-1 expression in primary tumor isolates has been positively associated with tumor size, her-2/neu, and the presence of extramammary metastases (4). ER is inversely associated with these same predictors of advanced disease, because ER-negative tumors are more commonly associated with her-2/neu and aggressive cancer (53). GPER-1 expression has been similarly correlated with the progression of serous ovarian adenocarcinoma (12) and with endometrial cancer (13), although a role for ER in these cancers is less certain. Distinct estrogen actions through GPER-1 is further suggested by the finding that ER antagonists function as GPER agonists, and these agents have been widely used in conjunction with GPER-1-selective agents to confirm estrogen-mediated physiology or pathology independent of ER.

Estrogen provides well-described renoprotective effects (54), and recent data have linked estrogen to GPER-1-dependent actions that are directed at renal tubules. Using microdissected renal tubule segments and isolated intercalated cells, 17β-estradiol, G-1, and the ER antagonist ICI 182,780 have been shown to promote intracellular calcium signals that are not detected in similar explant and cell cultures isolated from GPER-1-deleted mice (55). Additional evidence shows that GPER-1 action in the kidney may attenuate salt-dependent renal damage in female mRen2.Lewis mice independently of changes of systolic blood pressure that are measured in this genetic model (56). Specifically, Chappell and co-workers (56) showed that G-1 administration restored megalin expression in proximal tubules and reduced renal hypertrophy and proteinuria coincident with reduced measurements of the lipid peroxidation product, 4-hydroxynonenal, and urinary levels of 8-isoprotane. These data imply that GPER-1 action may provide protection from salt-induced renal injury by activating megalin-dependent protein reabsorption and reducing these established measures of oxidative stress. Based upon the inverse relationship between oxidative stress and EGFR in patients suffering from chronic kidney injury (57), these findings may suggest that estrogen-mediated transactivation of the EGFR by GPER-1 may provide protection from injury-associated oxidative stress.

GPER-1 is expressed throughout the central nervous system with highest levels of expression in the hypothalamic-pituitary axis, hippocampal formation, and brain stem autonomic nuclei (58). Rapid estrogen-mediated actions associated with GPER-1 have been measured in hypothalamic (59, 60) and hippocampal neurons (61, 62). Recent evidence has suggested that acute administration of the nonclassical ER agonists, diphenylacrylamide estrogenic compound or G-1, in a global ischemia model delayed the loss of hippocampal neurons (63), suggesting that GPER-1 may serve a neuroprotective effect. Collectively, these observations from breast cancer and from reno- and neuroprotective studies may suggest that GPER-1 plays an important role in promoting estrogen-mediated cellular survival responses.

Subcellular Localization and Endocytosis of GPER-1 and Its Potential Activation by Other Steroids

The fact that GPER-1 was initially isolated in the absence of knowledge of its ligand and that man-made selective ligands have been widely applied for discerning its function are valid reasons for scrutinizing the nature of its physiological ligand(s). Recent independent lines of investigation evaluating the desensitization mechanism of GPER-1 and its potential role in mediating aldosterone action indicate that the nature and the specificity of naturally occurring ligands for this receptor should be further investigated.

GPER-1 exhibits many of the expected characteristics of a plasma membrane ER and employs plasma membrane-associated enzymes to promote its action. However, immunohistochemical analysis using GPER-1 peptide antibodies has shown that GPER-1 is predominately distributed intracellularly in normal mammary epithelia and in intraductal and invasive mammary carcinoma (4). GPER-1 is similarly distributed in normal ovarian epithelia and in ovarian and endometrial cancers (12, 13). This distribution pattern is not atypical for GPCR, because intracellular accumulation is commonly detected in immunohistochemical analyses, e.g. as observed for the 5HT2A serotonin receptor (60, 64), adenosine 2A receptor (65), dopamine receptors (66), and cytokine receptors (67). Intracellular retention may reflect slow egress during receptor biosynthesis as a consequence of the multiple regulatory events in the endoplasmic reticulum and Golgi network as well as receptor down-modulation during reuptake. Collectively, these mechanisms act to limit the concentration of the receptor on the surface and the threshold of the sensitivity to its cognate ligand and active signaling. However, intracellular distribution of GPER-1 is not the rule (Fig. 2), because different groups have shown that GPER-1 is strongly localized to the plasma membrane of uterine epithelia (68), myometrium (32), renal epithelia (56, 69), hippocampal neurons (61), and fish oocytes (24, 35). Some discrepancies have been reported regarding the localization of GPER-1 to apical (56) or basolateral (69) surfaces of renal epithelia, an observation that is interesting considering the polarized functional activities of renal epithelia. These observations indicate that GPER-1 is not merely confined to intracellular membranes and may also suggest that individual tissues in the body modulate estrogen action via GPER-1 in response to intrinsic (gender, age) and extrinsic (available estrogen or other extacellular stimuli, injury) factors that determine its relative abundance in the plasma membrane.

Fig. 2.

Subcellular distribution of GPER-1. A, Recombinant HA-GPER-1 is expressed at the cell surface. Cells have been labeled at 4 C with anti-HA rabbit antibodies and then fixed, permeabilized, and immunostained with mouse GPER-1 antibody. Surface (red) and intracellular (green) HA-GPER-1 protein is detected using species-specific antibodies. B, Endogenous GPER-1 is detected at the plasma membrane of zebrafish oocytes using rabbit GPER-1 antibodies (green) reproduced from Pang and Thomas (81), with permission. C, Archival human biopsy tissue specimen stained with rabbit GPER-1 antibodies shows that the receptor is largely intracellular.

Ectopic expression of GPER-1 has yielded conflicting results with regards to its preferred subcellular localization. GFP-GPER-1 results in its predominant intracellular accumulation in the endoplasmic reticulum with little detectable receptor on the surface (5, 61). In contrast, GPER-1 and GPER-1 tagged with small viral-derived epitopes (hemagglutinin and fluagglutinin) translocates to the plasma membrane (5, 24, 27, 47, 61, 70, 71). One likely explanation for intracellular retention of GFP-GPER-1 is that the addition of the relatively large GFP domain (∼26 kDa of foreign peptide) may result in an appropriately folded receptor that is retained intracellularly, whereas the smaller viral peptide tags may be better tolerated by GPER-1 for appropriate posttranslational processing, folding, and export to the plasma membrane. In this regard, it is important to point out that GPCR are poorly expressed in their endogenous setting and due to their hydrophobic nature are not easily characterized by standard SDS-PAGE analysis. In addition, they are subjected to extensive carbohydrate processing during their biogenesis and export to the plasma membrane (72). Immunoblot analysis of recombinant GPER-1 indicates that multiple species are detected (47, 71). The predominant species of recombinant HA-GPER-1 detected with GPER-1 peptide antibodies has an apparent molecular mass of approximately 44 kDA with lesser amounts of molecular mass species near 60 and 80 kDA. Although the exact identity of these larger molecular mass forms is unknown at present, both are eliminated after removal of N-linked glycan chains with glycosidases, and the predominant 44-kDA band shifts to 38 kDA, which is the estimated molecular mass of the unmodified core protein (47). The 60-kDA bands are expressed on the cell surface as determined by antibody labeling experiments (71). Future experiments must further characterize the nature of these GPER-1 protein species.

Further insight regarding the subcellular distribution of GPER-1 has been gained by studying its intracellular trafficking from the plasma membrane (47, 61, 69–71). It is generally appreciated that upon binding their cognate ligands, activated GPCR undergo a series of adaptive changes that effectively prevent excessive receptor signaling and typically involve receptor phosphorylation, the subsequent recruitment of arrestins, and the internalization of the receptor by endocytosis. Because GPCR are often poorly expressed, the bulk of what is known about GPCR desensitization and endocytosis comes from studies that rely upon the tracking of recombinant cell surface receptors in HEK-293 cells by antibody labeling with amino-terminally positioned viral epitopes. Using this approach, the endocytic fate of GPER-1 has been compared with GPCR that are known to undergo recycling (beta-1-adrenergic receptor) or lysosomal down-modulation (C-X-C chemokine receptor type 4). GPER-1 has a short half-life on the plasma membrane (<30 min) and undergoes an unusual endocytic trafficking pattern that does not involve arrestins and results in receptor degradation in the 26S-proteasome (71). Using the approach of surface labeling with antibody, it was also shown that GPER-1 endocytosis is constitutive, because ligand is not required for entry into cell. Constitutive GPER-1 endocytosis also has been shown in an independent study by Sandén et al. (70) using antibody labeling. This result may raise questions regarding the ligand specificity of GPER-1. However, it is important to note that 17β-estradiol-dependent internalization of unlabeled GPER-1 expressed in HeLa cells has been demonstrated (61). Moreover, constitutive endocytosis of antibody-labeled GPCR is commonly, but not universally, reported, because more than 60 GPCR exhibit different degrees of constitutive endocytosis by this assay (73). This result has been interpreted to suggest that the energy necessary for a GPCR to adopt an active conformation required for receptor internalization is reduced in the presence of antibody (73). Future experiments addressing the endocytic fate of GPER-1 by the use of alternative labeling strategies or in alternative cell models may further identify the cellular events that serve to regulate the amount of available GPER-1 on the cell surface and therefore determine its sensitivity to 17β-estradiol, and possibly, other ligands.

Recently, GPER-1 has been evaluated as a mineralocorticoid receptor-independent mechanism for promoting rapid aldosterone action. Feldman and co-workers (74) have postulated that an aldosterone-GPER-1 interaction may stimulate vascular contraction via erk-1/-2 signaling. This tenet opposes data gained from a number of studies that show G1/17β-estradiol-dependent vasodilation in aortic and mesenteric preparations from various species (reviewed in Ref. 75). Endothelial removal or nitric oxide synthase blockade attenuates, but does not eliminate, G1/17β-estradiol vasodilation, suggesting a smooth muscle effect, which is consistent with the expression of GPER-1 in the medial layer (76). Moreover, the GPER-1-selective antagonist, G15, abolishes this G1/17β-estradiol action. However, aldosterone exhibits a vasoconstrictive effect on endothelial-denuded mesenteric vessels (77). In addition, chronic G1 administration lowers blood pressure in estrogen-depleted rats that show an activated renin-angiotensin system and increased aldosterone (78) or is normotensive in a salt-sensitive model (55). Finally, the studies by Feldman et al. (79) primarily involving isolated aortic smooth muscle cells with forced overexpression of GPER-1 showed a modest aldosterone-mediated, GPER-1-dependent increase in phosphorylated erk-1/-2 (<2-fold), but it is not clear whether this effect reflects the in vivo role of GPER-1 in the vasculature. Moreover, it is not clear from the published data whether aldosterone-mediated GPER-1-dependent erk-1/-2 activation is EGFR-dependent or linked to cAMP. On the basis of these results, it was proposed that GPER-1 also serves as an aldosterone receptor and therefore provides a possible mechanism by which aldosterone and estrogen may act in an opposing fashion on smooth muscle cells to regulate vascular function (79). Yet the concept that GPER-1 may serve as receptor for aldosterone is interesting in light of the fact that GPER-1 maps to the short arm of chromosome 7, at 7p22.3, in close proximity to a genetic locus that defines familial hyperaldosteronism II (FAH II) (80), which is associated with measurable increases in serum aldosterone and failure to respond to dexamethasone. It is intriguing to postulate that deletion or mutation of GPER-1 may account for these altered endocrine responses that are measured in patients with FAH II. The converse concept that elevated aldosterone measured in FAH II may trigger GPER-1 action is difficult to associate with the genetic linkage map suggested by the Lafferty et al. study (80). More formal development of the concept that GPER-1 serves as an alternative aldosterone receptor capable of eliciting pregenomic signaling requires the direct demonstration of its association with specific binding of aldosterone and its antagonists in ectopic and endogenous settings as well as testing the idea directly that aldosterone and 17β-estradiol (and other estrogens) compete for GPER-1 binding and signaling activity.

Conclusion and Future Directions

The body of published data supports the concept that GPER-1 acts independently and functions as a Gs-coupled plasma membrane-associated receptor that has specific binding activity for naturally occurring and man-made estrogens including: the primary female sex steroid hormones, estradiol and estriol, ER antagonists, phytoestrogens, and xenoestrogens. It shows no specific binding (or signaling) for the inactive stereoisomer, 17α-estradiol, cortisol, progesterone, or testosterone. Via 17β-estradiol, GPER-1 stimulates adenylyl cyclase and triggers the release of membrane-tethered EGF. GPER-1 expression is positively associated with prognostic variables that predict human disease progression in breast, ovarian, and endometrial carcinoma. Man-made selective GPER-1 ligands are useful tools that discriminate GPER-1 and ER and have been used widely to probe the physiological and pathophysiological effects that are associated with GPER-1, and these agents suggest a role in the central nervous system, in the kidney, and in vascular tissue. Efforts to better understand the mechanism of GPER-1 action and to determine the cellular factors that dictate its activity will complement these studies and lead to a better appreciation of the role of this newly appreciated ER in human disease.